From 2008 to 2011, to investigate the influence of host plants on the spread of TSWV, samples of crops and weeds were collected. To investigate natural host plants from plants with different life cycles, plant samples were collected in the winter (November to February), spring (March to May), and summer (June to October). Sampling was conducted in five areas where occurrence of TSWV has been reported continuously (Suncheon, Naju, Yeonggwang, Seosan, and Yesan) in South Korea (Fig. 1). Sample collection points were inside and outside of non-heated greenhouses that had been used as nurseries for hot pepper seedlings, and other crop cultivating greenhouses or open fields around hot pepper cultivating greenhouses or open fields.

Total RNA was extracted from collected samples with easy-spin Total RNA Extraction Kit (Intron Biotechnology, Seongnam, Korea) according to the manufacturer’s manual. TSWV infection was confirmed by RT-PCR with AMV Reverse Transcriptase (for cDNA synthesis, Promega, Madison, WI, USA), GoTaq DNA Polymerase (for PCR, Promega), and a TSWV-specific primer set (TSWV-6F, 5′-GAGATTCTCAGAATTCCCAGT-3′; TSWV-6R, 5′-AGAGCAATCGTGTCAATTTTATTC-3′), as described in a previous study (Tables 1,-3) (Ko et al., 2013).

For 4 years, to identify natural hosts of crops and weeds that can be infected with TSWV, a total of 1,104 samples was collected from weeds (830 samples) and crops (274 samples) from five areas in Korea where TSWV has occurred (Table 1, Fig. 2). Because weeds have various life cycles (summer annual, winter annual, and perennial), it is hard to find many different kinds of weeds if sampling is conducted at a certain time only, so sample collections were done repeatedly in the different seasons. In addition, we sampled in the open fields, heated greenhouses, and non-heated greenhouses as the weed species that can be identified can vary depending on heating during winter. Collected samples were confirmed as belonging to 144 species (34 crops and 110 weeds) (Tables 1,-3, Fig. 2).

Among crops, plants that belong to the family Solanaceae including tomato, eggplant, and sweet pepper showed high infection rates (75-100%) (Table 2). In the family Liliaceae, on the other hand, infection was observed only in shallot (Allium × proliferum). According to our observation in Yesan, shallot was cultivated in a non-heated greenhouse during winter as a perennial crop and was confirmed as an important host for propagation of western flower thrips (F. occidentalis). Thus, TSWV-infected shallot is a natural host for TSWV overwintering and propagation of the insect vector. In the family Fabaceae, three species (pea, peanut, and red bean) were confirmed as natural hosts in some places collected from an open field in Yeonggwang near many chrysanthemum cultivating greenhouses. From those farms, TSWV was also detected from chrysanthemum (19%), and large colonies of western flower thrips were observed. Based on this result, the TSWV insect vector (western flower thrips), which cannot overwinter in open fields (Brødsgaard, 1993), could directly damage the crops in the open field the next spring after overwintering in Chrysanthemum greenhouses. Tomato, eggplant (S. melongena), lettuce (Lactuca sativa), and crown daisy (Glebionis coronaria) were infected as seedling stages of the crops in nurseries with TSWV-infected hot pepper plants in Seosan.

Our results show that cooperation of thrips controls at neighboring farms is very important for effective TSWV control. In particular, the occurrence of the thrips in crops other than pepper, known as the most important host of TSWV, can be overlooked, but it is important because it can cause the spread of the thrips. In Korea, especially when farming scale is small, many kinds of crops are mixed and grown in other fields. As our results show, not only solanaceous crops but also beans, cucurbits, sesame, shallot, and curled mallow can be natural hosts of TSWV in South Korea. Since these are commonly grown crops in Korea, even if the pepper or tomato farmers manage thrips, it is very likely that viruliferous thrips including the western flower thrips (F. occidentalis) will continuously inflow from neighboring farms they are growing. This is also one of the reasons why TSWV problems continue to occur in Korea.

Table 3 shows infection rates of TSWV in weeds. Among the weeds collected in areas of TSWV occurrence (110 species), 42 species in 22 families were identified as TSWV-infected. Asteraceae, Brassicaceae, and Caryophyllaceae families had the largest numbers of infected species (6 species each). According to life cycle analyses of TSWV weed hosts, 18 species were confirmed as winter annual (42.9%), 30.9% (13 species) were summer annual, and 26.2% (11 species) were perennial (Table 3, Fig. 3). Among the weeds analyzed from numerous (>20) samples, most samples of summer annual Eclipta alba (Fig. 2G) showed TSWV infection (95.6%), and half of winter annual Stellaria media (55%) and perennial S. aquatica (54.5%) (Fig. 2C) samples were confirmed as TSWV infected. A larger number of winter annual weeds were confirmed as many of those species overwintered in heated greenhouses in TSWV occurrence areas. These results suggest that winter annual weeds play important roles in overwintering of TSWV. From our previous study on overwintering of Tomato yellow leaf curl virus (TYLCV), it has been shown that S. aquatica can act as a viral source to newly introduced insect vectors in the next cultivation after overwintering in the non-heated greenhouse (Kil et al., 2015). We did not confirm overwintering of TSWV in wintering S. aquatica plants by time-course follow-up and artificial experiments shown in the previous study on TYLCV, but it can be inferred that S. aquatica plants infected with TSWV play a similar role in overwintering and spread. S. media showed the highest infectivity among winter annual weeds, is the dominant species in greenhouses in the winter season, and had a vital role as a reservoir host for TSWV and a main host for western flower thrips. Perennial weeds including Artemisia princeps, S. aquatica, Calystegia sepium (Fig. 2F), and Phytolacca americana are common in cultivation areas in Korea, so it is likely that they can act as green bridges for TSWV.

TSWV occurs repeatedly through complex interactions between main host plants, other crop and weed hosts, and the insect vector thrips in Korea (Fig. 4). In hot pepper cultivating farms and nurseries, TSWV occurrence is common in Korea. The western flower thrips has spread to all areas of Korea. As shown in the above results, weeds growing in a cultivation area and other crops grown in the same or neighboring greenhouses also can be infected with TSWV and weeds act as intermediate hosts. In a nursery for hot pepper seedlings in spring, TSWV infection was detected from Cetastium glomeratum, S. aquatica, S. media, Lamium amplexicaule, and Galium spurium, which are winter annual or perennial weeds. If healthy plants are moved to an open field or greenhouse for cultivation, some viruliferous insect vector can infect them, creating additional inoculum for another vector. To overcome this cycle, control of insect vectors and weeds is necessary, and selection of TSWV-resistant cultivars can reduce economic damage. However, there have been a few reports for TSWV isolates that overcome Tsw gene-mediated resistance in pepper (Boiteux and de Ávila, 1994; Jiang et al., 2017; Riley et al., 2011).

TSWV has been a long-standing problem in over 100 countries, and its host range is the largest (Cho et al., 1986, 1987; Parrella et al., 2003). In Korea, TSWV has been indigenous as a representative plant virus that continuously damages pepper cultivation, and it is damaging not only peppers but also cultivation of tomatoes, potatoes and chrysanthemums. In addition to the known hosts, new hosts are also being reported in Korea. H. japonicas, one of weeds that are easily observed in farms, was reported as TSWV-infected (Yoon et al., 2018a). Soybean (Glycine max) cultivated commonly and famous ornamental plant P. obtusifolia introduced from Taiwan and China have also been identified as TSWV recently (Yoon et al., 2018b, 2019). These reports mean that TSWV host range analyses for the commonly cultivated or newly introduced crops, and the weeds found on farms should be done continuously to update lists of TSWV hosts, which can be helpful to design strategy for effective controls. This study was conducted to investigate crop and weed hosts of TSWV from areas near hot pepper cultivating farms in Korea and 23 crop species and 42 weed species were confirmed as TSWV-infected in fields. Our updated list of natural hosts from this study can be used for TSWV control.